dures are represented as simple lists with items to complete. These checklists are available for each transition to the next phase and each change within a phase. Additional explanations are available for people using the list for the first time. See figure 1.

quired to know how to write database entry files, and the table can be used as a document of the registered variables. Adding the Acnet interface to the LabVIEW program typically takes less than a day unless an intrinsic control is required using a custom-made console application. Often though, there is no need for a specific console application and the generic ready-to- use parameter page suffices.
Any major change that the developers make must be docu- mented according to the commissioning change list unless the change is irrelevant to the operators or programmers of the console application. This judgment is left to the developers to avoid too much paper work while an error at this stage would not halt accelerator operations. Any significant change must be noted in the IB and the version number of the software must be updated. The version number is also available as a LabVIEW variable and can be seen on-line on a console for easy reference with an IB entry.
To bring the system up to operational status, the developer must provide the means to rebuild the system from the ground up, complete the IB entry, complete the spare parts list, pass the system certification test, and notify the Operations liaison. The system certification includes a review of the analysis and a complete from-the-ground-up software installation on a differ- ent computer to prove that, in case the computer fails, it can be resurrected on a spare.

APPLICATIONS
So far, six projects have been created using the platform. All are hooked up by a tokenring card to the control system and can be remotely controlled over ethernet.
Synchrotron Light Monitor
The transverse image of the beam formed by synchrotron radiation is displayed on a Quadra 950. The light is converted by a CCD camera to video signals which are read into the Mac by a frame grabber card. A stepper motor card interfaces to the positioning motors which aligns the camera within the tele- scopic system. A GPIB card controls a CAMAC crate which has timing and power supply modules. The image is captured and analyzed by a LabVIEW add-on package. All hardware and software were commercially available and needed only minor modification. See [2].
Beam Line Tuner
On injection of the beam in the ring, the turn-by-turn beam positions are measured to estimate the initial phase and ampli- tude of the betatron oscillation. A console application page re- trieves the results and calculates the beam line corrections. The Macintosh IIci is hooked up to a VXI crate through the MXI interface. The VXI crate contains VXI digitizers, cards for de- coding of injection events and beamsync signals, and RF-mul- tiplexers to switch between beam position signals. See [3]. Sampled Bunch Display
A Quadra 950 controls two GPIB scopes taking samples of the proton and pbar bunches in the Tevatron and analyzes their

Installation list

To-operational list

Operational change list

Figure 1. The development phases.

To start a project, the developer obtains the project folder which contains the lists and the documentation about the Acnet interface and Macintosh utilities to be used. As the pro- ject progresses, all notes and created documentation should be added to the folder. The first list tells which utilities should be installed so that every instrumentation computer has the same (trusted) utilities and everyone can expect the same tools to be available. Because the development process in the stand-alone phase is a continuous series of changes and additions, too many to document, there is no list for each change but a re- quirement to backup the software at least once a week.
To continue to the commissioning phase, the list for the transition must be completed. List items require notifying the involved groups, (at least the Operations Group), backing up software, and adding an entry in the Instrumentation Book (IB). An entry in the IB includes the status of the project, a diagram of the configuration, a description of its operation, the table of LabVIEW variables accessible by the control system, a description of diagnostic procedures, and operator training sheets. The IB is located in the Main Control Room and is in- tended for use by the operators. A software tool is available to help integrate the application to the control system. It gener- ates from the same table that LabVIEW uses to initialize its side of the interface, the database file that the control system needs to register the variables accessible in LabVIEW. In this way, both sides use the same table, the developer is not re-

beam intensities. Timing signals come from a GPIB controlled CAMAC decoder card. A major part of the program is the control logic to set the scope and analysis for the various states of the accelerator. The intensities are displayed in the lab wide status display and used for luminosity calculations. See [4].
Ion Profile Monitor
For both vertical and horizontal planes, a 30 point beam profile from micro-channel charge pickups is sampled. All 20,000 turns of the Booster cycle are captured by VME digitiz- ers. CAMAC modules provide timing signals and control of the high voltage power supply of the pickups through a GPIB interface. The analysis determines the position and emittance for each turn. The data, 2.5 Mbytes, can be displayed for any turn individually or as a whole cycle, (decimated for display), in a color intensity plot. The Quadra 650 takes about 15 sec- onds to retrieve the data and display the results. See [5]. Collision Point Monitor
Each side of the collision region has a horizontal and verti- cal pair of beam position pickups. A GPIB multiplexer directs the signals from beam position pickup plates to a GPIB scope. The LabVIEW program retrieves the data and determines the position and time of a proton and pbar bunch to derive the col- lision point. The results are accessed through a standard pa- rameter page on a control console.
Flying Wires
Transverse beam profiles are obtained by flying wires through the beam and digitizing the resulting beam loss inten- sities and current wire positions. A Macintosh Quadra 950 controls the wires' movement using a Nubus motor controller card. In a VME crate, controlled through a MXI interface, digi- tizers sample the beam loss signals on each turn of the beam. The LabVIEW program reads the data in and analyzes the pro- files to determine the emittance of the beam. Work is in progress to speed up analysis and create a console application to present the results to the operators.

EXPERIENCES
Because LabVIEW is a graphical language, one is saved the time normally spent debugging syntax errors. Additional time is saved by the availability of many drivers for the data acquisi- tion hardware. Very little time is spent on the display or communication at the Macintosh side due to the built-in graphics. If, however, the generic parameter page does not suf- fice, a console application must be written, (in C or Fortran), which can take extra time. Thus by using available hardware and the LabVIEW software, we found that most time, up to eighty percent, is spent on programming the analysis routines and the control flow of the program.
Besides the availability of LabVIEW for the Macintosh other useful utilities are available as well, e.g., screen snap- shot takers, installer programs, and remote control programs.

A disadvantage of the Macintosh is the lack of a real-time kernel. The main factor that alleviates this lack is that our ap- plications are geared towards monitoring, not controlling. Results are presented to the operators who don't require a fixed time interval in which an answer must be ready. Time-critical operations, such as timing the data acquisition, can all be done in hardware. The LabVIEW program only checks to see if the data acquisition is complete and if the device must be repro- grammed to change the measurement setup. It is possible to do time-critical operations if one could assure that nothing else than the LabVIEW program would be running on the Macintosh. However, we use remote control programs like Timbuktu and PlanetX. These can be used anytime and inter- rupt a time-critical operation.
We found that system crashes are often due to hardware in- compatibilities, e.g., accelerator cards and expansion chassis, and software incompatibilities, e.g., system extensions. The solution was to replace or not use the offending parts. Crashes due to program errors are normally found and corrected for in the stand-alone or commissioning phase. To handle the crashes at remote locations, the Macs can be rebooted by cycling the power using a remote controlled switch. The most common reason to reboot a machine is after an unsuccessful change in the LabVIEW program or because an upgrade of a utility proved incompatible with the particular Mac. To date the longest uninterrupted operation of a Mac was several months.
The procedures for the development and maintenance pro- cess are new and we just started applying them to the projects. We expect to make modifications as needed.
All developers had a favorable impression of the LabVIEW- based platform and estimated that it would be much harder and more time-consuming to implement a similar application on an ASCII language based system.

ACKNOWLEDGMENTS
I like to thank the many people in the Instrumentation, Controls, and Operations Groups for their advice and help.

REFERENCES
W. Blokland, "An Interface From LabVIEW To The Accelerator Controls Network", Proc. of Accelerator Inst. Workshop, Berkeley, USA, 1992, pp. 320-329. A. A. Hahn and P. Hurh, "Results From An Imaging Beam Monitor In The Tevatron Using Synchrotron Light", HEACC'92, Hamburg, Germany, July 1992, pp. 248-250. W. Blokland, "A VXI/LabVIEW-based Beamline Tuner", PAC'93, Washington.
E. Barsotti, "A longitudinal Bunch Monitoring System using LabVIEW and high-speed oscilloscopes", To be published at 1994 Accelerator Instrumentation Workshop in Vancouver. J. Zagel, "Booster Ion Profile Monitor using LabVIEW", To be published at 1994 Accelerator Instrumentation Workshop in Vancouver.

[1]

[2]

[3]

[4]

[5]